xorshift | Implementations of a couple of xorshift PRNGs

I am trying to port the "RayTracing in One Weekend" into metal compute shader. I encounter this strip artifacts in my project:

Is it because my random generator does not work well?

Does anyone have a clue?

I would like to implement the following C program in Perl:

I'm trying to create a pseudo-random generator API, but numbers generated by xorshift have unrandom nature. You can see the algorithm and tests here:

I'm trying to pass the output of a function in C to itself a number of times but I'm still getting the same answer each time (I tried to pass it by value and by reference I do not know which is the correct one). I have two functions in my program one that does the calculations and it is called xorshift() and the second function has a for loop; to call the function multiple times and it is called foo(). What I want to do exactly is :

I want to pass (f) to the xorshift function and do the xor and shift operations and then return the output, and this output should go to xorshift function and do the xor and shift operations again and get the output and send it again to xorshift function ..etc I want to do this operation for a number of times.

Any thoughts or suggestions on how to do it?

Follow-up question for IEEE 754 conformant sqrt() implementation for double type.

Context: Need to implement IEEE 754 conformant sqrtf() taking into account the following HW restrictions and usage limitations:

1. Provides a special instruction qseed.f to get an approximation of the reciprocal of the square root (the accuracy of the result is no less than 6.75 bits, and therefore always within ±1% of the accurate result).

2. Single precision FP:

   a. Support by HW (SP FPU): has support;

   b. Support by SW (library): has support;

   c. Support of subnormal numbers: no support (FLT_HAS_SUBNORM is 0).

3. Double precision FP:

   a. Support by HW (DP FPU): no support;

   b. Support by SW (library): has support;

   c. Support of subnormal numbers: no support (DBL_HAS_SUBNORM is 0).

I've found one presentation by John Harrison and ended up with this implementation (note that here qseed.f is replaced by rsqrtf()):

I need a high performance random number generator that is thread-safe. I need only random bytes in the value type (which is ulong for now), not within ranges. I've used the C# built-in Random class, but it was kind of slow and not thread-safe.

Later I moved to XORShift functions that actually works very fine, but to achieve thread-safeness I need to put the calculation in lock, and that degrades the performance drastically.

What I'm using to generate a random ulong is the following:

I am reading Java Concurrency in Practice and encounter the following code snippet. Listing 12.5 https://jcip.net/listings/PutTakeTest.java

I'm trying to make a game in C fit under a 3KB limit, and to do so I want to remove all standard library references. The only header I want to have in my program is windows.h and so far I'm doing pretty good.

The only external references I have are one call each of rand(), srand(), time(), and getch(). This may seem like a lot for a game under 3KB, but the time, rand, and srand functions are only so that when the game starts it can have a random seed.

rand to get the numbers, srand to set the seed, and time to get random seeds. So if I could find a way to get rid of the excessive random generation procedure, I could get rid of 3/4 std library functions. I'm not so sure what to do about getch though, but ill focus on that later.

Some people have reccomended xorshifts, but it seems like that requires the type uint32_t, which belongs to stdint.h. I also tried using integers instead, and while the resulting number was random, it was random in the sense that the number given back was random, yet you could depend on it giving you that random number each time. I could give it time as a sort of seed, but then id still have to use the time function. Is there something I'm overlooking, or should I use a different method?

EDIT: as per request, my compilation flags are -ffunction-sections, -fdata-sections, -s and -Os. my only linker flags are -Wl,--gc-sections and I am not sure what dynamically linked and statically linked libraries are completely, but I have a decent idea. My code as is comes out to 12KB, and if i use UPX i can get it down to 7KB. my code is here:

I am reading Java Concurrency in Practice and encounter the following code snippet.